How Is Rock Coring Different from Soil Boring?

Rock coring is the continuation of a soil investigation that begins once the borehole reaches rock. A normal soil boring takes samples and performs the SPT (Standard Penetration Test, ASTM D1586) at intervals of about 1.0–1.5 m, but when the stratum becomes so hard that the SPT sampler can no longer be driven (called Refusal), the bit has reached rock or very hard soil.

This is a key decision point. If the project needs deeper data — for example high-rise foundations, piles that must be socketed into rock (Rock Socket), dams or tunnels — the drilling crew "switches mode" from soil sampling to cutting continuous rock cores, because the rock can no longer be assessed by the SPT N-value; it must be judged from the actual rock material, its fractures and its strength instead. (For the overall number and depth of boreholes, see the article How many boreholes and how deep?)

What Equipment Is Used for Rock Coring? — NMLC Core Barrel and Diamond Bit

The main tool for rock coring is the core barrel. SPN uses the NMLC type, a double-tube core barrel that produces rock cores about 52 mm in diameter. The outer tube rotates to cut the rock, while the inner tube stays stationary to support the core so it does not twist or break under the rotation.

In hard rock, the barrel must be fitted with a diamond bit to cut into the rock, because diamond is harder than almost every rock-forming mineral. During drilling, cooling water is pumped down the drill rods to dissipate heat, reduce bit wear and flush rock cuttings out of the hole.

  • NMLC core barrel (double tube) — recovers ~52 mm cores with minimal sample disturbance
  • Diamond bit — cuts hard rock continuously with a consistent cutting edge
  • Cooling water — controls temperature and flushes cuttings during drilling
  • Core box — holds the cores in depth order for transport to the laboratory

The model below shows the site when the borehole reaches rock: the rotary drill rig switches to a double-tube barrel with a diamond bit, and the recovered cores are laid in the core box in depth order. The right panel is a section through the NMLC double-tube barrel.

3D model of rock coring with an NMLC double-tube core barrel, diamond bit and rock core box The left image is a 3D site model. A rotary drill rig stands over a borehole at the edge of the cut-away section. The borehole passes through fill, soft clay and stiff clay to about 14 m, then enters limestone. At the bottom of the hole is the core barrel with a rock core inside. On the ground is a core box with the cores laid in rows in depth order, with a label for each run's depth interval. The upper cores near the rock surface are broken into shorter pieces than the lower ones. The right image is a section through the NMLC double-tube barrel: the outer tube rotates with the diamond bit to cut the rock, while the inner tube stays stationary on a bearing to support the core so it does not twist or break. A core spring holds the core at the bottom end, and cooling water flows down the gap between the tubes to cool the bit. The core is about 52 mm in diameter. 3D model — Rock coring Not to scale Rotary drill rig Switch to core barrel Core box In depth order, run labels Top of rock ~14 m SPT refusal Core run ~1.0–1.5 m Core in barrel Top zone fractured More intact deeper Depth compressed · schematic NMLC double tube (section) Bearing Inner: no spin Outer rotates drives the bit Inner static holds the core Core ~52 mm Cooling water flows in annulus Core spring Core Lifter Diamond bit Diamond Bit
3D model — rock coring continues from the soil borehole once it reaches the top of rock (about 14 m in this example). Cores are laid in the core box in depth order; the upper part near the rock surface is usually more fractured than the deeper part. Right panel: the outer tube rotates the diamond bit to cut the rock, while the inner tube stays stationary to support the ~52 mm core, with cooling water flowing down between the tubes.

How Is Rock Core Sampling Carried Out?

Rock core sampling to ASTM D2113 is done in intervals called core runs, each about 1.0–1.5 m long, as follows:

  1. Drill through the soil to the top of rock, then switch to the NMLC barrel with a diamond bit.
  2. Core the rock one run at a time, controlling thrust, rotation speed and cooling water to suit the rock type.
  3. Pull the barrel out of the hole and carefully remove the core from the inner tube without breaking it further.
  4. Lay the cores in a core box in depth order, labeled with the project, borehole number and the depth interval of each run.
  5. Measure the Core Recovery and RQD of each run and record them in the coring log together with rock type, weathering grade and fracture characteristics.
  6. Transport the core boxes to the laboratory with care and store them for compressive strength and other rock property tests.

Careful handling is essential: if cores break in transit, an RQD re-measured in the laboratory will be lower than the true value, and specimens to be trimmed for compressive strength testing may end up shorter than the standard requires.

How Are Core Recovery and RQD Calculated, and How Do They Differ?

Core Recovery (CR) and RQD (Rock Quality Designation) are the first two numbers an engineer reads from rock coring results. Both are measured directly from the cores of each core run, but they answer different questions — CR answers "how completely was the sample recovered?", while RQD answers "how good is the rock that was recovered?"

CR = (total length of core recovered ÷ length of core run) × 100%Core Recovery — shows how completely the sample was recovered
RQD = (Σ length of intact pieces ≥ 10 cm ÷ length of core run) × 100%Rock Quality Designation per Deere's method — counts only pieces broken along natural fractures; breaks caused by drilling or handling are not counted

🧮 Example: a 1.50 m core run recovers 1.50 m of core in total → CR = 100%. But the intact pieces 10 cm or longer add up to only 0.70 m → RQD = 0.70 ÷ 1.50 × 100 ≈ 47%. This means the rock was fully recovered, but it is closely fractured and its quality is rated Poor.

Comparison Table: Core Recovery (CR) vs RQD

← Scroll to see the full table →

Core Recovery (CR) vs RQD
CriteriaFormulaRangeInterpretation
Core Recovery (CR)(Total length of core recovered ÷ length of core run) × 100%0–100%Measures sampling completeness — a low value may indicate highly weathered rock, voids, or core lost during drilling
RQD (Rock Quality Designation)(Σ length of intact pieces ≥ 10 cm ÷ length of core run) × 100%0–100% (5 classes per Deere)Measures rock mass quality — the higher the value, the fewer the natural fractures and the more intact the rock

An important point is that RQD can never exceed CR, because it counts only intact pieces, which are a subset of the recovered core. So if CR is high but RQD is low (such as the CR 100% / RQD ≈ 47% example above), the rock was recovered almost completely but is broken into many short pieces.

How to calculate Core Recovery and RQD from the core of one core run, per Deere's method The core from one 1,200 mm core run is laid out from top to bottom. The first piece is intact and 250 mm long. The second piece is 200 mm long but highly weathered and fails the soundness criterion, so it is not counted in RQD. The third section totals 250 mm, broken into fragments shorter than 100 mm and highly weathered, so it is not counted. The fourth piece is intact, 190 mm. The fifth piece is 60 mm, shorter than 100 mm, so it is not counted. The next pieces are 80 and 120 mm, separated by a drilling-induced break, so they are counted together as one 200 mm piece. The bottom 50 mm was not recovered. Core Recovery equals 1,150 divided by 1,200, about 96 percent. RQD equals 250 plus 190 plus 200, which is 640, divided by 1,200, about 53 percent, rated Fair. The values in the figure are an example to illustrate the calculation. Core Recovery vs RQD — from one core run Not to scale 1 core run = 1,200 mm CR: all pieces RQD: intact pieces ≥ 100 mm only 250 250 Intact 200 0 Highly weathered, fails 250 0 Pieces < 100 + weathered 190 190 Intact 60 0 < 100 mm 80 + 120 200 80 + 120 drilling break 0 0 Lost mm mm Core Recovery (CR) Total core recovered ÷ run (250+200+250+190+60+80+120) = 1,150 ÷ 1,200 mm ≈ 96% RQD (Deere) Σ intact ≥ 100 mm ÷ run (250+190+200) = 640 ÷ 1,200 ≈ 53% RQD counting rules Intact piece ≥ 100 mm → count Shorter than 100 mm → no Highly weathered, unsound → no Drilling break (red dashed) → join pieces, then count RQD ≤ CR always 53% = Fair · example value
Measuring Core Recovery and RQD from the core of one 1,200 mm core run per Deere's method — CR counts every recovered piece (≈ 96%), while RQD counts only intact pieces ≥ 100 mm long and joins pieces separated by drilling breaks (≈ 53%). The values in the figure are an example to illustrate the calculation.

What Does RQD Tell You About Rock Quality?

RQD is used to classify rock mass quality according to Deere's criteria, the standard classification commonly used in site investigation reports.

← Scroll to see the full table →

Rock mass quality classification by RQD (Deere)
RQD (%)Rock mass qualityEngineering meaning
0–25Very PoorHeavily fractured rock with closely spaced fractures; behaves close to hard soil or weathered rock
25–50PoorMany fractures; rock mass strength must be assessed with caution
50–75FairModerately fractured rock; common in the upper part near the rock surface
75–90GoodFairly intact rock with widely spaced fractures
90–100ExcellentIntact rock with almost no fractures; rock mass strength approaches that of the intact rock material

Besides RQD, the coring log also records the weathering grade of the rock, from Fresh to Completely Weathered, and the hardness of the rock material; these are used together to interpret the condition of the rock strata.

Classification criteria used in the coring log: rock quality by RQD, weathering grade and rock material hardness A three-column diagram. The first column is rock mass quality by RQD per Deere: less than 25 percent Very Poor, 25 to 50 Poor, 50 to 75 Fair, 75 to 90 Good, and 90 to 100 Excellent. The second column is six weathering grades: G1 Fresh, G2 Slightly Weathered, G3 Moderately Weathered, G4 Highly Weathered, G5 Completely Weathered, and G6 Residual Soil. The third column is seven rock material hardness grades R0 to R6, from extremely weak to extremely strong, with approximate uniaxial compressive strength ranges per ISRM guidelines. Coring log criteria — RQD · Weathering · Hardness Not to scale RQD (%) Rock mass quality · Deere Weathering Degree of Weathering Rock hardness Hardness · UCS ~ISRM < 25 Very poor Very Poor 25–50 Poor Poor 50–75 Medium Fair 75–90 Good Good 90–100 Excellent Excellent G1 Fresh Fresh G2 Slightly Slightly G3 Moderately Moderately G4 Highly Highly G5 Completely Completely G6 Residual soil Residual Soil R0 Extremely soft UCS ~0.25–1 MPa R1 Very soft UCS ~1–5 MPa R2 Soft UCS ~5–25 MPa R3 Medium strong UCS ~25–50 MPa R4 Stiff UCS ~50–100 MPa R5 Very stiff UCS ~100–250 MPa R6 Extremely strong UCS > 250 MPa
Classification criteria used in the coring log — rock mass quality by RQD (Deere), weathering grades G1–G6 and rock material hardness R0–R6, with approximate uniaxial compressive strength ranges per ISRM guidelines

Case Study: Rock Coring in Suphan Buri Province

An example from an actual SPN project (May 2026): borehole SPT-BH-2 in Suphan Buri Province was drilled through soil from the ground surface to about 14 m with normal sampling and SPT testing. On reaching limestone, the crew switched to rock coring with an NMLC barrel and diamond bit, continuing to a depth of 20 m.

Coring from 14–20 m gave a Core Recovery of 100% in every run, while RQD ranged from 32–79%: the upper part near the rock surface was closely fractured (Poor to Fair quality) and improved with depth. This is exactly the kind of data engineers use to assess the depth at which the rock becomes sound enough to carry foundation loads. (To learn how to read a full borehole report, see the Boring Log reading guide.)

Boring log for borehole SPT-BH-2, a project in Suphan Buri Province, showing soil from 0–14 m and limestone from 14 m down where the work switched to rock coring
Boring log for borehole SPT-BH-2 — soil from 0–14 m tested by normal SPT before limestone was encountered (Image: SPN Soil Engineering)
Coring log recording limestone core samples at 14–20 m depth, showing Core Recovery of 100 percent and RQD of 32 to 79 percent, a project in Suphan Buri Province
Coring log for the limestone from 14–20 m — CR 100% in every run, RQD 32–79% improving with depth (Image: SPN Soil Engineering)

What Laboratory Tests Are Done on Rock Cores?

Well-preserved rock cores are screened, and intact pieces are trimmed and tested in the laboratory. The main tests are:

  • Uniaxial Compressive Strength (UCS) to ASTM D7012 — the compressive strength of the rock material, used as the key parameter for assessing the bearing capacity of the rock stratum
  • Point Load Index to ASTM D5731 — a quick test used to estimate compressive strength when specimens are too short for UCS testing
  • Unit weight and water absorption — basic properties that support rock classification

Combined with RQD and the fracture data from the field, these test results give the design engineer the information needed to assess rock mass strength for foundation design. The appropriate values for each project must be determined case by case by the engineer responsible for the design.

Summary

Rock coring is an extension of soil investigation for projects that need data on the rock strata. An NMLC barrel fitted with a diamond bit cuts into hard rock, the cores are laid in core boxes, Core Recovery and RQD are measured in the field, and the cores are then carefully transported for compressive strength testing in the laboratory. All results are summarized in an investigation report certified by a licensed civil engineer (COE license), which can be used for foundation design and submitted to the relevant authorities. (For an overview of the service and starting prices, see Soil investigation prices.)

📌 Note

This article is provided as engineering knowledge. The drilling depth, number of boreholes and interpretation of rock quality for each project should be decided at the discretion of a licensed engineer.